|
Post by Colin Comard (Delay) on Sept 29, 2008 20:51:17 GMT -6
[Can someone add a question mark to the title?]
This is mostly aimed at Cat and Satoshi, so yeah.
I'm making another Appendix (apart from the above mentioned) in my soundfont tutorial, part 3, "Tips from the professionals" and it will feature tips from you guys. It can be about anything, from the pitch a certain instrument should be to how many instruments (layers) should be in a present.
And I will give due credit.
Edit: To give me a tip, you should have released at least 1 soundfont.
|
|
|
Post by Colin Comard (Delay) on Oct 11, 2008 21:56:47 GMT -6
oh i will. but this would be more like something I heard from satoshi a caouple times:
"Use more than 1 layer [fer each present] so that the sound is louder"
That sort of thing. The arranging splits is goin to be a video appendix. To avoid making people read scrolling text.
|
|
|
Post by satoshilyish on Oct 11, 2008 23:01:15 GMT -6
Whoops, I didn't even notice this thread. Let me think..I definitely use more layers in order to make things louder..I keep it usually at a 2 layer limit for most instruments (aside from the heart, which I make 3 since the bass tone has to overcome all the other melodies) and my animal chorus (whigch I made 20 in slightly out of sync/pitch/phase, etc) For making loop points in Soundfonts, the obvious tip is to zoom in all the way, and make sure the beginnign and end pieces match. Also, very important, use a spectrum analyzer for the segment that you're looping to make sure what you highlighted is the correct pitch. Had such a headache getting that worked out. To work around the 5 note limit you can attach layers from other sound samples. Example is what I did in my Zelda Orchestra piece. I did it with the hearts (made it a plane sound with boats) Another important part. If possible, save your master files as the highest levl dls file. I ran into a problem where SF2 files don't save the volume and pitch adjustments directly on the sound samples itself, but clears that info and saves it on the layers. Its kind of a pain when you have to make valume adjustments and have to go to EVERY SINGLE LAYER to make them. Alternatively, the DLS files are capped at 2 for the "key number to pitch" option, which would make things annoying if one of your instruments (say the mushroom or boat) had higher values than that (SF2 doesn't have this limitation), so its a trade off of convenience between the two formats. I'll probably think of other tips later on.
|
|
|
Post by satoshilyish on Oct 12, 2008 13:49:48 GMT -6
Another thing I remembered. I sometimes use the built in tuner in AWave so the sound files are properly tuned , but there are some flaws with this. I don't know about Viena (need to mess with it more) but you can't select specific spots in order to tune. Generally its an average of the frequencies of the sound file, which won't make it perfectly tune. Either you have to find the longest duration of a fixed pitch in the sound file and set that, or set the closest to the beginning sample to the right pitch (so rapid firing the notes doesn't sound off).
Also, the tuning mechanism tunes according to what the loudest lowest frequency is, in that order. This was why I was having so much trouble with tuning the swans, because by default it would tune to one of the chord frequencies, and not the actual base frequency since it was softer than the next frequency up, and I had to attempt tuning with my own ear (which obviously didn't go so well). I ended up downsampling the file until the frequency that I knew was correct, read correctly in [Cool Edit's] tuning feature, and from there I was able to manually tune it exactly within 1 cent.
lol, I know most of this probably sounds very confusing to the average person, so I apologize for my technical babble.
|
|
|
Post by AdolfoBaez on Oct 18, 2008 21:16:39 GMT -6
*Edit: To give me a tip, you should have released at least 1 soundfont. ..
I´ve made a lot of soundfonts on Vienna, I haven´t released any coz´there are some pretty good sf available already.
|
|
|
Post by Colin Comard (Delay) on Oct 19, 2008 11:30:56 GMT -6
Oh yeah, and every sound sample you use on Viena, must be on C (unless you plan to transpose), or else the note you put on C in whatever program you're using the soundfont in will not[/i] be a C[/quote] middle C right?
|
|
|
Post by Colin Comard (Delay) on Oct 20, 2008 17:56:41 GMT -6
right. Unfortunately, sometimes it's hard to make it the right sound you want AND have it be C. But thanks for the tip, i'll add it.
|
|
|
Post by satoshilyish on Oct 20, 2008 18:43:05 GMT -6
Oh yeah, and every sound sample you use on Viena, must be on C (unless you plan to transpose), or else the note you put on C in whatever program you're using the soundfont in will not[/i] be a C[/quote] What? That sounds kind of dumb. Although, you can simply resample down to the right pitch, right?...right? (Awave already looks to be a lot better in that regard)
|
|
|
Post by cypher8301 on Nov 22, 2008 2:59:58 GMT -6
In Vienna, if you create a sample and do not assign a root key it defaults at C4 (key 60). Its pretty easy to pick out by ear what you need though, and assign it the proper root key. So any sample should not be a problem, just make sure to have a healthy sample for more complex instruments or sound will start getting weird once you get further out from your sample base.
Cypher
|
|